In one sense, this is odd. Standard economic theory says that people want to smooth their spending over time. This implies that when bad times come, people should reduce their saving and increase their borrowing in order to maintain their spending habits, so the savings ratio should fall in recessions which should help to mitigate recessions.
So why hasn’t it done so? One obvious reason is that credit constraints prevent some people borrowing and make others save for fear they’ll not be able to borrow in future. Another reason is that expectations for future incomes have been downgraded, and this naturally should reduce spending now.
One fact, however, alerts us to the possibility that something else is going on. This is that this recession is not unusual in seeing the savings ratio rise. It also rose in the 1981 and 1991 recessions.
One aspect of this something comes from a study of the effects of the Dutch postcode lottery, in which winners get a new BMW. Researchers have found that when someone wins the BMW, his neighbours are more likely to buy a new car in the following months. This suggests that our spending really is motivated in part by a desire to keep up with the Joneses.
A second aspect comes from an experiment in a restaurant in Market Bosworth. Researchers varied the background music between pop, classical and nothing. They found that diners spent significantly more when the classical music was played. This, they say, is because such music creates an upmarket atmosphere which encourages more spending.
These two findings suggest that consumer spending is more cyclical than conventional smoothing theory predicts. In good times, such as the mid-00s, our neighbours’ high spending and the general atmosphere of prosperity encourage us to spend more. But in bad times, these influences go into reverse and we spend less.
This implies that, far from being a stabilizing force for the macroeconomy, consumer behaviour can be a destabilizing one. This is partly disguised by the fact that a big portion of our spending is on essentials such as food and electricity. However, if we strip these out, we see that it is actually more volatile than GDP generally. Since 1997Q1, the standard deviation of quarterly real consumption growth excluding food, housing and utilities has been 1.1 percentage points, compared to just 0.7 percentage points for GDP growth.
Does this destabilizing matter? Herein lies a paradox that doesn’t get the attention it deserves. On the one hand, most macroeconomists assume that it does; this is why they devote their careers to considering how best to use monetary and fiscal policy to stabilize the economy.
On the other hand, however, voters seem relaxed about macroeconomic fluctuations, as there’s no demand for better insurance against such swings; voters are hostile to raising unemployment benefits, and Robert Shiller’s proposal for macro markets to insure us against economic fluctuations never caught on. This corroborates Robert Lucas’s suggestion that the welfare costs of macroeconomic instability are small.
Whether fluctuations matter or not, though, my point is the same – that consumer behaviour has played more of a role in this downturn than generally realized, and this might be for reasons which are not wholly consistent with conventional theories of consumer spending.
